types, anyone? :)
Vladimir Komendantsky
Would you be interested to hear and see me describing the typed
approach to functional programming at clojure dojo?
I'd be happy to have a few sessions dedicated to these accessible
introductions by my former boss:
http://hal.inria.fr/inria-00329572/en/
http://hal.inria.fr/inria-00001173/en/
The first one aims at proving correctness of simple imperative
programs and might be a good starting point.
In fact, I thought to start with is a very good compiler verification
course available online:
http://cristal.inria.fr/~xleroy/courses/Eugene-2010/ . But it might
seem a bit obscure at first approach. Looks more like a possible goal
provided everyone has seen the above papers.
Best wishes,
Vladimir
Rick Moynihan
Sounds interesting Vladimir, I've been recently reading "To disect a
mockingbird (A Graphical Notation for the Lambda Calculus)", as I
begin to explore more of the theoretical underpinings of FP.
I'm sure others would, like me, like a gentle introduction to the more
academic side of the FP discipline.
However, I'm unlikely to be in Dundee much longer, as I've been
offered some contract work in Manchester. This said, you might want
to contact Barry Carr (or Andy Gibson) to see if you can get a talk
through the Scottish Developers.
Thanks again for the offer and the links... I'll maybe try and digg
through those papers at some point... if they're not too heavy for me,
that is! :-)
R.
p.s. There was recently a discussion on the Clojure mailing list about
type checking in Racket/Clojure etc... It might be of interest:
http://groups.google.com/group/clojure/browse_thread/thread/f0e419142915847d#
Vladimir Komendantsky
It's a pity you are leaving Dundee... I mean, congratulations on your
new job. ;)
Thanks for the link to the discussion thread. I think Clojure
community has yet to discover the long-standing dualism between typed
and untyped functional programming. In the thread, someone cited a
paper by Matthias Felleisen, a famous guy in the FP community, and
probably his student. It's a good exposition of what would Scheme
become should it have types. Well, they may have an implementation
already...
Best wishes,
Vladimir